RESEARCH PAPER NUMBER THIRTY-SEVEN
RELATION OF DISPERSION
SPECIAL CEMENTS „ '. '
EDW. VV. SCRIPTURE, JR., PH. D.
Master Builders Research Laboratories
CLEVELAND, OTU« \
COPtfSJGHT l^V"' THE ;M;A€Ur GUILDERS COMPANY
APPLICATION of the principle of dispersion to
Portland cement has received widespread recogni-
i ion by the construction industry in large defense
construction as well as other building. In previous
ers this principle has been described in its relation
in normal portland cement and the economy of its
application to concrete mixes has been shown.
1 ! always several means of accomplishing
ed result and it is no more than reasonable to
select that which is most effective and economical. In
concrete to-day, the special cements offer means of
securing certain specific properties desirable for
ticular structures and it seems appropriate to consider
their relation to cement dispersion. This paper has
been prepared to give the construction industry in-
formation on this subject.
Cement dispersion will accomplish many of the
objectives sought in the use of special cements. More-
over, it is not incompatible with them but, on the
contrary, is equally effective with the special cements
as with normal portland cement in improving the
properties of concrete. As a result it is found that^in
many cases, the desired properties can be secured with
a normal portland cement and a dispersing agent more
economically than with a special cement and without
jome of the disadvantages which the latter may have.
In other cases it is shown that the desired end can he
realized most effectively and economically by app
tion of the principle of dispersion to the special cement.
Believing that knowledge of these relations will be
of value to the construction industry whenever the use
of a special cement is contemplated for some specific
purpose, The Master Builders Company has published
This paper reviews briefly the mechanism whereby greater efficiency
lined through dispersion. The effect of this action on
the | es of concrete and the relation of cement dispersion to
economy are reviewed.
The various usual types of special cements and the particular
• tiance are described. These include:
High 1 Type III for high early
Type V for resistance to
1 1 and [V for reduction of heat
eat e\ olution and corro-
•• increased workability and
' Ified workability and
luction in capillary attraction.
dispersion is applicable to these
at ii is to normal Portland cement
• ame effect on the special
ed for a given consistency and in
or hydratioi i titly
Liability, watertight i time
i rsion and sp-
me relation to tl
e properties of tl
RELATION OF DISPERSION TO SPECIAL CEMENTS
Table of Contents
Mechanism of cement dispersion 6
Effect of dispersion on properties of concrete 7
Economics of cement dispersion 7
Types of Cement
High Early Strength cements 10
( lorrosion Resistant cements .14
Low Heat cements .17
Pozzolan cements 20
Natural cement blends 22
( Vments with grinding aids 26
Waterproof cements 30
RELATION OF DISPERSION TO SPECIAL CEMENTS
THE mechanism whereby dispersion of portland cement profoundly
affects the properties of concrete has been described in a previous
paper (Research Paper No. 35). In another paper (Research
Paper No. 36) the economic aspects of this principle have been dis-
cussed, specifically in comparison with addition of extra cement to the
mix. This material is briefly reviewed in the following three sections.
Nature of Cement Dispersion
Application of the principle of dispersion to hydraulic cements has
recently assumed considerable importance in concrete construction and
has found widespread use in the present defense building program.
Dispersion of finely divided solid material in a liquid is not in itself
new as it has been employed in ceramics, dyeing, and other fields.
Dispersing agents are, however, specific in nature so that a reagent
which acts to disperse some particular solid-liquid system may or may
not act in a similar manner in some other system. Furthermore, an
effective dispersing agent may interfere with other reactions of the
system. The use of dispersion in the field of concrete and mortar has
awaited the development of effective cement dispersing agents which
would not interfere with the normal reactions of hydraulic cements.
Such reagents were discovered about 10 years ago and since that time
have been developed in the laboratory and in the field so that they are
now established as a practicable means of considerable importance of
improving the properties of concrete.
When portland cement is mixed with water the individual particles
tend to gather together and stick to each other in clumps, i.e.. the
sohd-hquid system is flocculated. This is due to lack of mutually
Cement Suspended in Water Cement Suspended in W T ater
repellent electrostatic charges on the cement particles. If a suitable
dispersing agent is introduced into the mix the clumps are broken up
and the cement then acts as individual particles, i.e., is dispersed.
cf. Fig. I.)
Dispersion of the cement produces two important effects. The
water which had been trapped within the particle clumps is released to
become a part of the mixing or placing water. The surface area of the
cement in contact with water is greatly increased since the particles
are no longer in contact with each other. As a result of the first the
amount of water required in the mix for a given consistency is less, i.e.,
the water-cement ratio is reduced. Since the value of the cement is
dependent on a hydration reaction which is a surface phenomenon, the
second effect which promotes more rapid and more complete hydration
permits more efficient utilization of the cement. By the reduction in
water-cement ratio and by the increase in surface area of cement
available for hydration the potential value of the cement is more
Effect of Cement Dispersion on the Properties of Concrete
Those properties of concrete which are dependent on the surface
area of the cement and on water-cement ratio, and this includes most
of them, must necessarily be improved by dispersion. These effects are
realized in the concrete in both its plastic stage and subsequent to
During the plastic stage dispersion of the cement in a given mix
will produce more placeable concrete with less water due to release of
water from the cement clumps. The fattiness of the mix is increased,
while segregation and bleeding are reduced, due largely to the increased
effective surface area of the cement. Volume change before hardening
is markedly decreased due in part to the lower water content and in
part to the greater surface area.
A greater uniformity and freedom from gross defects of the hardened
concrete may be expected from the improved properties in the plastic
stage. Greater watertightness with reduced permeability and absorp-
tion are realized through the lower water content required for placing.
Higher strengths and very greatly increased durability with respect to
both freezing and thawing and sulphate corrosion may be attributed to
the lower water-cement ratio of the dispersed mix and to the increased
surface area available for hydration.
It is not suggested that cement dispersion is a panacea for all ills.
Poor concrete will continue to exist due to poor workmanship, poor
mix design, defective materials or other causes whether the cement
used is or is not dispersed. What cement dispersion will do is improve
the quality of good concrete or minimize the defects of poor concrete
by attacking the problem along the fundamental line of using the
cement more effectively.
Economics of Cement Dispersion
Granted that the application of the principle of dispersion to cement
will produce definitely large improvements in the properties of the
concrete the question will naturally arise whether the value of the
improvement is greater than the cost of dispersion. The economies to
be effected will lie in the original cost of the materials making up the
concrete mix, in the initial cost of the concrete in place as affected by
workability, segregation, etc., and in the eventual cost of the structure
as determined by maintenance and length of life of the structure.
The last two of these are somewhat intangible and difficult to
express in specific monetary terms. There can be little question that a
more placeable concrete with less tendency to segregation will reduce
labor costs in placing, finishing, patching, and rubbing. \\ hether these
savings in themselves will pay for the cost of dispersion will depend on
the conditions on any particular job. Likewise it is hardly to be doubted
that a more watertight durable concrete will reduce the cost of main-
tenance and effect a saving by extending the life of the structure.
Whether the savings so effected justify the cost of dispersion will depend
on the severity of the conditions to which the structure is exposed.
Material costs can be compared much more definitely and are of
general applicability. A suitable basis of comparison can be found in
the relation between cement dispersion and the amount of additional
cement required to produce similar results. Although there will be
minor variations in different localities the cost of dispersion in an
average concrete mix of between five and six sacks of cement per cubic-
yard may be taken approximately to equal the cost of adding three-
quarters of a sack of extra cement.
Data have been accumulated both in the laboratory and in the
field which show that dispersion of the cement in a given mix will pro-
duce a strength equal to or greater than one additional sack of cement,
will produce greater workability with respect to both mobility and
cohesiveness, greater durability with respect to resistance to both
freezing and thawing and sulphate corrosion, and will increase water-
tightness with respect to both absorption and permeability to a greater
extent than will one additional sack of cement. Further, dispersion of
the cement will reduce volume change and will decrease heat evolution;
entirely beneficial results whereas additional cement will have the
In view of these results it is evident that cement dispersion is a
more economical means of producing a desired effect than is the addi-
tion of extra cement. How this economic advantage will be utilized
will depend on the requirements in each case. When high quality is not
a factor it will be used to produce a given quality of concrete at a
lower cost. Where quality is the primary consideration dispersion will
produce maximum quality at a given cost.
It is generally recognized that there are many other methods of
modifying one or more properties of concrete than that of additional
cement and it would seem appropriate to consider the relation of dis-
persion to these.
The properties of concrete are influenced both by the nature of the
materials from which it is made and the proportions in which they are
assembled or the mix design. It is not proposed to deal with the purely
physical aspects of mix proportions or characteristics of the aggregates.
It is assumed that advantage will have been taken of these factors
within the limitations of available materials. It is proposed to consider
here the benefits and disadvantages which are derived from some
modification of the cement itself, that is, the use of special cements.
TYPES OF CEMENT
Portland cement is composed of a number of compounds, chiefly
of lime, produced by burning together calcareous and silicious materials
and grinding the resultant clinker to a certain degree of fineness. The
more important chemical compounds of portland cement are, tricalcium
silicate (CaS), dicalcium silicate (C2S) and tricalcium aluminate (C.3A).
These constitute the bulk of the cement and more is known concerning
their influence on the properties of the cement than is known con-
cerning a considerable number of other compounds present in relatively
It has been realized for a long time that the properties of portland
cement could be materially altered by modifications in its process of
manufacture. This has been given recognition by the recent adoption
by the A.S.T.M. of five official classifications of portland cement. The
principle factors influencing the properties of the cement are:
1. Its compound composition which in turn is influenced by the
nature of the raw materials and the burning process.
2. The fineness to which the cement is ground, i.e., its surface area.
3. Additions of more or less reactive materials to the cement,
usually during grinding.
The rate of cooling of cement clinker has an influence on the prop-
erties of the cement but this may be included as part of the burning
The five types of cement recognized by A.S.T.M. standards are
Type I. Normal portland cement
Type II. Moderate heat of hydration cement
Type III. High Early strength cement
Type IV. Low heat of hydration cement
Type V. Corrosion resistant cement
Cements falling under any of these types may be further modified
by additions of less than V f of extraneous materials, usually grinding
aids, under certain conditions. There are also various other types of
cements such as blended cements, masonry cements, natural cements,
etc. Additions of extraneous materials other than those recognized by
the A.S.T.M. have also been used. The various modified cements
discussed in this paper are as follows:
1. High early strength cements
2. Corrosion resistant cements
3. Moderate or low heat of hydration cements
4. Pozzolan (blended) cements
5. Natural cements
6. Cements with grinding aids
7. Waterproof cements
ould be pointed out that the reason for the development of the
special or modified cements has been the feeling that while normal
Portland cement has proved and still is eminently satisfactory for most
purposes, in certain applications one or more properties in a higher
degree than they would be present in normal portland cement would be
advantageous. This has meant, usually, that in the special cement,
one property has been markedly enhanced without much change in or
at the expense of the otb
I. High Early Strength Cement
As its name implies, an high early strength cement is one in which
lgth develops rapidly. This effect mav be produced in three ways
1 by altering the compound composition, increasing CaS and possibly
I A. at the expense o- - 2 finer grinding to produce Greater surface
area, and 3 the addition of an accelerator.
The advantages of high early strength are well known and need no
more than passing mention. They include more rapid rate of con-
struction more frequent reuse of forms, reduced curing costs, and
earlier utilization of the structure.
The disadvantages of high early strength cements are that usually
a higher water-cement ratio is required for a given consistency and a
more rapid rate of heat evolution produces greater temperature rises in
concrete. The setting characteristics of the cement mav also be altered
so that they may affect finishing operations adversely. This la*t
however, does not appear to be necessarily the case
TABLE No. I
High Early Strength cement and Normal Portland cement with
Dispersing Agent Mod.
Concrete Mix -Cement
3 1 1 inches
High Early Strength Normal Portland Cement
Gallons water per cu. yd.
Lbs. per sq. in.
Average for three separate series.
with Dispersing Agent mod. I
\\ hether a high early strength cement should be used in any
specific case depends on the relative importance of speed and the other
properties of the concrete. The properties of concrete made from high
early strength cement, aside from the early development of strength
are probably not affected to any marked extent, except in the case of
large masses where temperature effects are of prime importance. On
the other hand they are not improved in comparison with a similar
concrete made from normal port land cement and properly cured. There
is some reason to believe that the slower development of strength of the
normal Portland cement has some beneficial effect on the ultimate
properties of concrete. Where speed is a primary requisite or is of pre-
dominant importance then the use of a high early strength cement
should be considered as one means of realizing this objective. It will be
considered, presumably, on the basis of economy, that is, whether it is
the least costly way of securing the desired strength at early ages.
Two other methods of producing high early strengths immediately
suggest themselves. These are the use of additional cement (normal
Portland) and dispersion. With respect to the former, sufficient addi-
tional cement will undoubtedly give strengths equal to those of high
early strength cement. The use of additional cement to accomplish this
purpose involves the disadvantages of higher volume change and heat
evolution. Furthermore it seems hardly possible that equal strengths
can be secured by the addition of an amount of portland cement equal
in monetary value to the increased cost of the high early strength
cement. Were this the case, high early strength cements would have
no economic justification.
I Z 3
,,11,'M^ 1 — — i "*
aK0 i«>*^ — — -*
, £ t C 1 _ -*— ■**
CEMENT 44-7 LBS.
SAND 1346 LBS.
STONE-*. 1988 LBS.
SLUMP_ 34 IN.
Normal portland- Dispersed 31 gals. ""
Age in Days
Dispersion of the cement will increase the strengths at all ages,
necessarily including the early ages. This is due first to the reduction
in water-cement ratio made possible by dispersion and second to the
higher surface area made available for hydration. With a slight
modification the use of a cement dispersing agent in the mix will give
with normal portland cement early strengths equal to those of high
early strength cement and higher strength at later ages (Table I and
Fig. ID. Likewise the dispersing agent will produce strengths at early
ages higher than one additional sack of cement (Fig. III).
1 n-r i I I | ! | I j lL ji
" COMPRiWVf STRf^TM AT fARL Y STAGES ^*
" Dtspfwm ft AWtrtoML etfflttr ^^ ^>_
rh -/ it
% t "V
5 2 «*5
5 ' ^v
% 1 -*t?
^ / "27
'§ j£ -S
£ M it ,-.-
£ %tjl ^— - —
^ lit $J =*"
1 it^-2 ^*.'
t t%f ^g^
X s &
X 1 7 jS^
1 iJ~ ^2
fe ti a
1 t y
1 -44 -f
i 11 Z
# =E ±
I Z 3 4- 3" 6 7
Age in Days
It should be noted that, although these relations between high early
strengths and normal portland cements hold in a general way, there are
considerable variations between different brands of cement. Thus, with
an high early strength cement which gives unusually high early strengths
and a normal portland which gives lower than average early strengths,
the latter used with the dispersing agent may fall somewhat below the
former. Conversely, with an high early strength cement of lower than
average and a normal portland with higher than average early strength,
the latter when dispersed may give considerably higher strengths than
the former. This is to say that the relative early strengths for the
high early strength cement and the dispersed portland cement in any
particular case will depend on the relation between the particular
brand of each which is being used, but in general, with average cements,
the early strengths will be approximately the same.
The economic relations of these three methods of securing early
strength can now be considered. Dispersion will cost approximately
$0.40 per cu. yd. for an average concrete mix. One additional sack of
cement will cost approximately $0.50 per cu. yd., and the use of a
high early strength cement instead of normal portland cement about
$0.70 per cu. yd. It is evident that the most economical way to secure
the desired early strength is by cement dispersion.
There are certain advantages, in comparison with the use of high
early strength cement, in securing the required strength by cement
dispersion. Since, by this means, the increases in early strengths are
secured in very large part by reduction in water-cement ratio the heat
evolution will be much less. The setting characteristics of the concrete
are not appreciably altered. Since dispersion permits a large reduction
in water-cement ratio the well-known advantage of lower water-cement
ratio such as increased durability, increased watertightness, and
decreased volume change are also realized.
It should not be overlooked, however, that there is no incompati-
bility between a suitable cement dispersing agent and an high early
strength cement. Actually, since the finer high early strength cements
show an even greater tendency toward flocculation than the coarser
normal cements, cement dispersion is even more effective. Therefore,
dispersion of the high early strength cement will produce still higher
early (and ultimate) strengths as well as the other advantages of lower
water-cement ratio. Strength curves for an high early strength cement,
dispersed and undispersed are given in Fig. IV.
IJ n tatSSEO
] T - U H0I$WZ°
CEMENT . 447 LBS.
SAND 1346 LBS.
STONE- 1 1988 LBS.
SLUMP 2| IN.
Undispersed Cement. 38 gals.
Dispersed Cement „ 33gals.
1 1 1 | | | | | | | | | 1 1 1 1 1
Age in Days
In this connection a comparison may be made, from an economic
point of view, between dispersion and approximately one additional
sack of high early strength cement. These will both give about the
same strength Fig. A' . Cement dispersion, for an average concrete
mix. costs about $0.40 per cu. yd. while the additional sack of high
early strength cement will cost about $0.63. The conclusion is inevit-
able" that the more economical way to secure high early strength is by
HIGH EARLY STRENGTH CEMENT
iz: - ;s*. sacks
337 LBS. ....CEMENT 44.7 LBS.
i 390 LBS SAND 1346 LBS.
2056 LBS STONE-f. ...1988 LBS.
33 GALS WATER 39 CAiS.
3 IN SLUMP. 2i IN.
• i 2 3 4 s e - fi 9 io :; : =
Age in Da
Three methods of securing higher early strengths have been con-
of high early strength cement, use of additional normal
and cement, and the dispersion of the cement normal portland .
— . the most economical is dispersion and this has the added
. : improved properties springing from a lower water-cement
ratio. When exceptionally high early strengths are required the most
omical method of realizing them is by the use of a cement di
with an high early strength cement.
II. Corrosion Resistant Cements
Corrosion resistant cements are designed to resist the efTec
corrosion, particularly by sulphate solutions including sea water. This
is accomplished primarily by the reduction of the ntent of the
cement as this is the compound most readily attacked by sulpha
A comparison of a normal portland cement Type 1 with a coma
resistant cement Type V is shown in Fig. VL
Reduction of the CiA content of the cement reduces the rate and
amount of heat evolution which would appear to be a desirable effect.
It also tends to retard setting and rate of hardening which is not un-
desirable except where speed of construction is a factor. Apparently the
other properties of the concrete are not materially affected.
Where resistance to corrosion is important, as in structures exposed
to sea water, to sulphate bearing ground waters, and to highly polluted
atmospheres there would appear to be good reason for the use of
corrosion resistant cements. There is no economic reason why, under
such circumstances, corrosion resistant cements should not be used
as these are usually obtainable at no increase in cost over normal
A. Normal portland cement
Compressive strength at 28 days — 4660 lbs./sq. in.
After 2 years in 8% Magnesium Sulphate (changed weekly)
B. Corrosion Resistant cement
Compressive strength at 28 days — 5200 lbs. sq. in.
After 2 U years in 8% Magnesium Sulphate (changed weekly)
Dispersion will increase resistance to corrosion because of reduced
permeability of the concrete, increased hydration under Riven curing
conditions/and improved structure. Additional cement will also in-
crease resistance to corrosion due to increased strength and lower
permeabilitv. The decrease in sulphate corrosion secured by dispersion
-•ater than that secured by addition of one extra sack of cement
Since the cost of dispersion is less than one additional sack
of cement the more economical \\; cure corrosion resistance with
normal portland cement is to design the mix with a dispersing agent.
Moreover the other advantages of dispersion compared with additional
cement with respect to strength, resistance to freezing and thawing,
watertightness, volume change, workability and heat evolution are also
Research Paper No. 36 .
Magnesium Sulphate kly)
use of a the
relative degrees of corrosion resistance of the two particular cements
selected as there are considerable variations within each type.
Where corrosion resistance especially is sought it would seem desir-
able to use a corrosion resistant cement of Type V if available. Cement
dispersion, however, is equally applicable to a Type V cement as it is
to a Type I cement because both are naturally flocculated. The same
advantages with respect to workability, strength, durability and other
properties, including resistance to corrosion are secured. Data on the
effect of a dispersing agent on a corrosion resistant cement are given
in Table II.
Stone — %"
Lbs. per sq
- 28 days
Here it will be seen that dispersion has similar effects with this type
of cement to those which it has with a normal Portland cement. Con-
sequently, for a given degree of corrosion resistance it will be most
economical to design the mix with a Type V cement and a cement
dispersing agent or for a given cost the highest resistance to corrosion
can be secured in this manner.
III. Low and Moderate Heat of Hydration Cements
Cements with low heats of hydration were developed to circumvent
the deleterious effects of rapid rate and total amount of heat evolution
in raising the temperature of the concrete, particularly in mass con-
crete. There is some reason to believe that these heat effects are also
of more importance in thinner sections than has usually been believed 1 .
The reductions in heat evolution are secured by reduction of the C:jA
content, or increase in the C2S content at the expense of the CaS, or both.
These cements have somewhat slower rates of setting and hardening
than normal port land cement. Except when speed of construction is a
factor this is probably beneficial rather than detrimental. Resistance
to sulphate corrosion is usually improved. It does not appear that the
other properties of the concrete are materially affected.
The difference between the low (Type IV) and the moderate
(Type II) heat of hardening cements is one of degree rather than of
kind. The low heat cement is not generally available, being supplied
only in special cases for large mass concrete projects. The moderate
heat of hardening cement is available in many markets and is frequently
the same as or interchangeable with the corrosion resistant cements
(Type V). These cements are usually supplied at no increase in cost
over that of normal portland cement so that there is no economic
reason why they should not be used where heat evolution is an im-
*L. A. Forbrich — J. A. C. I., Sept. 1941.
THE EFFECT OF
DISPERSION ON THE
/ / \
F A l\
THE EFFECT OF DISPERSION ON
THE EARLY RATE OF HEAT
LIBERATION OF A MODIFIED
LOW HEAT CEMENT (TYPE H)
f Jtf >^:^>
■^V^ _^^ |HOURS AFTER MIXING]
TABU] No. HI
Ileal of Solution
: I I
• I I
( londil ion
. per gram
7 days 28 d;
lar effecti wii li
if I) normal Portland i i
A further advantage of cement dispersion is found in the effect on
heat evolution. Dispersion of a low or moderate heat cement in a given
mix does not appreciably affect the rate or amount of heat evolution
(Table III and Fig. VIII). Within the limits of accuracy of the deter-
minations there are no differences in heat evolution for the dispersed
and undispersed conditions of each cement. Therefore, a mix with a
lower cement content and a lower rise in the temperature of the concrete
(Fig. IX) can be designed using a cement dispersing agent at a lower
cost, with equal or greater strength, and with equality or superiority
with respect to other properties.
— i— i 1 1 1 !— *
MOD F ED 10
™ ft +
70 ■- uj
60 w +
DO ■ r 52
*rr*/r iVi> /W K/7 ,
„ Or tOT*"'
WW T 1
.A 5 -JU
4° £ if
^ 11 A
FECT OF DISPERSION ON
"EMPERATURE CURVES FOR
^L MASS CONCRETE MADE
MODIFIED LOW HEAT CEMENT
Wl I H >
o I I I I
n i 1
1 AGE IN DAYS h
The situation also may arise where low heat evolution is desired but
rather higher strengths at the early ages are required than would be
secured with the low heat cements (Types II and IV). In this case it is
possible to take advantage of the fact that dispersion, with normal
Portland cement, will permit use of a minimum cement content for the
required strength. Since the heat evolution is not materially affected
by dispersion this means that the temperature rise in the concrete for a
given strength at the earlier ages is reduced as is illustrated in Fig. X.
n^ r-i i i ■ ■ ' i ! :
^NORMAL PORTLAND CEMENT (TYPE I) f
risers***?. ^Zmcnt />£*
„fl K "
\ / J^
THE EFFECT OF DISPERSION
ON THE COMPUTED ADIABATIC,
TIME-TEMPERATURE CURVES FOR
TYPICAL MASS CONCRETE MADE
WITH NORMAL PORTLAND CEMENT
30 ■ i
13 7 28
Where the heat evolution of the concrete is a primary consideration
a low or moderate heat of hardening cement should be selected. A mix
should then be designed, using a suitable dispersing agent, with mini-
mum cement factor consistent with the required strength and quality
of concrete w.'th respect to other qualities.
IV. Pozzolan Cements
A pozzolana is a material, usually of a silicious nature, which
will combine with lime in the presence of water at normal temperatures
form cementitious compounds. It may be of natural or synthetic
origin. When added to a Portland cement mix a pozzolanic material
will combine, more or less rapidly and completely, with the free lime
the cement. This reaction produces desirable results, particularly
with respect to corrosion resistance, on the properties of the concrete
but i he extent of the improvement effected varies widely with the
nature of the pozzolanic material.
From time to time special pozzolan cements have been made by
grinding together Portland cement clinker and a pozzolana. The
quality of the resultant cement was dependent on the nature of the
clinker, the nature of the pozzolana, and the proportions in which
they were inter-ground. Assuming that the nature of the materials and
their proportions are such that a satisfactory pozzolan cement will be
produced certain beneficial effects may be expected. These are a
greater resistance to corrosion (sulphate), (Table IV), lower and slower
heat evolution, and greater workability as far as cohesiveness or fatti-
ness is concerned (Table V). The disadvantages are slower setting and
slower rate of development of strength. They may also require some-
what higher water content for a given consistency than would a normal
TABLE No. IV
Resistance to 10% sodium sulphate of 1:5:6 concrete made with
fly ash 20% cements.*
Compressive Strength at 6 mo,
Fly Ash Sulphate Unexposed
None 6310 5560
No. 1 6950 7120
No. 2 6800 6580
No. 3 6650 6400
No. 4 6420 5830
No. 5 5480 5060
TABLE No. V
Settlement or Bleeding of Concrete before Hardening
Original level at 24 hours
t 400 g.
Gravel y s ' 450 g.
Slump (6" cone) 2 in.
specimens 2" x 4"
In their properties pozzolan cements are very similar to low or
moderate heat of hardening cements and their applicability would
presumably be the same. A possible greater "fattiness" of the pozzolan
cement might constitute some advantage for this type but a choice
between them and the low heat types would probably be based on
economic considerations. When suitable pozzolamc matenals are
plentifully available the pozzolan cements can probably be produced
at lower cost.
♦Raymond E. Davis — Properties of Cements and Concretes containing
Fly Ash, June, 1938.
produced from a naturally occurring "cement roc] h has a
composition similar to that of the Portland cement raw mix. The
cementitious qualities of the "rock" are developed by burning
temperatures somewhat lower than those used for pn
Portland cement clinker.
In certain respects natural cements are superior to normal Portland
cement although they vary widely among them one
natural cement imparts desirable pr<> does not
necessarily imply thai some other natural cement will do like
The advantageous proper! ies of rial oral 1 1 ree of
cohesiveness or "fattiness" which reduces bleeding and
greater durability, and lower hea ion. The chief disad
are slower rates of setting and hardening, lower strength, until
after very long curing, and a I iter req
There is no agreement on thi i natural
cements. It may be t hal I hi
the lower burning temporal are produc< I irface < i lition which fa
greater water retentivity and hei " the other hand, the
properties of those cemei en attributed to the menl
of a higher proportion of air in the mix which ; ribed
to small amounts of oil or lentally (or purpo luded in
the cements during their manufacture. es in
durability might be explains ' hich
would help to prevent surface >r to inci ntrained
More probably both influent i
tion, as also t lie slower set
caused by slower hydrat ion.
Natural cements h;r e fail
to Portland cement. A mix ! >m natural c< ould
have undesirable pro in that it would be
and would have insufficienl strength unless cured for a long time,
of the recent use of natural i i n for highwa
sumably with the objective of reducing bleeding and and
thereby preventing scald e road Fig. XI] A and B). Usually
a 15* o to 20* ! replacement of portland i I
been used, that is about 1 sa< k il mix. Tl
reasonable degree of fattiness and air entrainmei
portion of natural cement produces too mucl fatl
qualities and entrains too much air SO tl
concrete is impaired.
Dispersion Of portland cement will I the
properties of the concrete similar ired with natura
replacements. Fattiness or col
reduced. The greater fattiness mix will
air entrapment as a factor in durability. This particular prope
be enhanced by a slight modification of the
improvements, if secured througl
instead of by the use of natural
advantages of the latter. Dispersion pro
strengths at all ages, does not affect I
Dispersed and undispersed portland cement mixes and a natural
Portland blend subjected to freezing and thawing cycles using calcium
chloride for thawing on the top surface.
Left B. Portland cement concrete- dispersed.
Center — A. Portland cement concrete undispersed.
Right — C. Natural portland blend — undispersed.
itural portland cement blend concrete!
'i thawing cycles using calcium chloride for
rtland blend dispensed.
and blend und
and permits a lower water-cement ratio instead of a higher one cf.
Research Paper No. 35).
It may, therefore, be possible to secure with normal portland cement
and a dispersing agent those properties which are desirable in the
portland-natural blend, fattiness and durability, without the lower
strengths and slower hardening of the blend (Table VI). Whether
this is the case will depend to some extent on the particular natural
cement used, as these cements vary quite widely in their properties.
A recent study showed that of two natural cements one was very effec-
tive in improving durability whereas the other had no effect in this
An illustration of the production of similar properties to those
secured with natural cement by dispersion of portland cement is shown
in Fig. XII (B and C). Here are given photographs of two cement
blocks made under similar conditions to those encountered in road
building and finished as would be a road. Their condition after a
number of cycles of freezing and thawing by the application of calcium
chloride to the frozen surface shows the reduced bleeding, increased
durability and resistance to scaling of the dispersed cement mix.
A cement dispersing agent for portland cement is also a dispersing
agent for natural cement and will, therefore, produce the same effects
with a blend as with an all portland cement mix. It is, therefore,
possible to design a portland-natural blend concrete mix using a cement
dispersing agent which will show greater improvement in properties
than will the use of additional cement equal in monetary value to the cost
of the dispersing agent. This is illustrated by the strength data given
in Table VI for two concrete mixes, dispersed and undispersed, of
equal cost. The added resistance to scaling so secured is illustrated in
Fig. XIII and in Table VII.
Portland and Portland- Natural Cements with Dispersion
Blend-Portland 83 y 2 %
Cement — lbs. 574
Sand - lbs. 1134
Gravel -lbs. 2099
Water -gals. 31.2
Slump in. l'x
Lbs. per sq. in.
3 days 2630
7 days 2790
28 days 3430
Durability — Portland-Natural Cement Blend
I'ndispersed and Dispersed J
Natural Portland Cement — lbs. 373
Natural Cement — lbs. 94
Sand _ Jbs. 1276
Stone- _ lbs . 1992 1992
^ ater -gals. 36 29^
Slump in. 2 2
f " c lpss in weight after freezing and thawing
20 cycles 6.2 14
50 cycles 40.0 7.4
r^Ji?^ be pointed °"t that since dispersion and the use of natural
S b9 i h mcre ^ e *]* fattin , ess of the mix - «'hen a dispersing agent
desifr! ?th^i a v?. 0rtlan< i l " natUral C ? me . nt blen '' ( ' are shouki be taken to
moT d ffi,^ ™H aV ° ,d Kf CesSlve fattl ^ SS as this wou,d make finishing
more difficult and mig ht cause excessive a r entrainment This mav
be accomplished by avoiding an excessively rich mix and an excessively
large proportion of natural cement in the blend. ««*s« eij
The desirability of using portland-natural cement blends is still a
h^T^Vi ' Mal Subject even w " h «*P«* to highway work it
has gained little acceptance in other fields. Assuming however that
are Sesfred'lnd tV 7t Y^ tUra ' C6me , nt blend ' ^nc.l^fattinest
w?th n- tu rll rllnf i he T^f™" 1 "< P*rt of the portland cement
with natural cement is contemplated, cement dispersion offers a means
( " in ;i I^rtlan<l cement mix withouT^he
J5 - of a portland-natural cement blend ami at
equa or less cost. If, for some reason, the use of a Dortland-natiir-.l
Me. then .he mix can be lefl t , ro ^
-' economically wit 1, a cement disjersmg agent
VI. Cements with Grinding Aids
-rials have been used as grradine aids that is to
nK ; ne,s or surface area. Among mSSTAS
The Vinsol cement suffers from the same disadvantage as a port-
land-natural blend in that strengths are lower. This decrease in strength
appears to be greater for the Vinsol resin cement than for the blend
and, moreover, it is permanent, that is, with long curing the strength of
the concrete made with a portland-natural blend will probably equal
that of a straight portland mix whereas the Vinsol resin cement will
always show lower than normal strengths. Apparently the loss in
strength is directly related to the amount of air entrained. Some strength
data on cements ground with and without Vinsol resin are given in
Table VIII. It is especially to be noted that, although the water-
cement ratio is lower for the Vinsol resin cement, due to the replace-
ment of water by entrained air, the strengths are markedly lower
instead of higher as would be expected.
TABLE No. VIII
Vinsol Resin Ground Cement
x — Cement
- Stone -
. per sq.
7 days 28 days
2 '• ,
2 ; (
13! i ti
It has been observed that the behavior of Vinsol resin cements is
quite variable. The effect of a given percentage of Vinsol resin, as
measured for example by the decrease in unit weight of the concrete
which is an index of the amount of air entrained, seems to vary with
different cement clinkers; even with clinker from the same mill pro-
duced at different times. This variation implies that the effectiveness
of the Vinsol resin with respect to air entrapment, bleeding, segrega-
tion, fattiness, durability, and strength will vary. In the present state
of our knowledge these variations are unpredictable. This means that,
because a certain percentage of Vinsol resin is ground with the cement
clinker, it does not necessarily follow that the desired properties will
be secured in the same degree. A better criterion of the effectiveness
of a Vinsol resin cement than the percentage used would seem to be the
drop in unit weight of concrete made with the cement in question
compared with a similar mix from normal portland cement. To find
this weight loss, however, seems to be a matter of trial and error from
one lot of cement to another.
Vinsol resin cements have been offered and to some extent used as a
substitute for a portland-natural cement blend, to produce similar
properties in the concrete mix, particularly increased fattiness and less
tendency towards scaling. They are competitive on a cost basis as
neither adds substantially to the cost of the concrete. Considering the
variable performance of the Yinsol resin cement and the losses in
strength which may result, it would seem that if the quality of work-
ability which either Yinsol resin cement or the portland-natural blend
will impart is desired, the choice would be on the latter.
A cement dispersing agent will disperse a Yinsol resin cement just
as it will any other type of cement and consequently will have the same
beneficial effect on the properties of concrete made from this variety of
cement. The use of a dispersing agent with a Yinsol resin cement,
however, is not recommended because the variability of these cements
presents the danger that the dispersing agent with the Yinsol resin may,
on occasion, produce excessive fattiness and air entrainment.
Just as in the case of natural cement blends it may be possible to
produce with normal portland cement and a dispersing agent those
qualities of fattiness and durability which it is intended to secure with
a Yinsol resin cement (Fig. XI Y). Likewise these advantages are
obtained without the lower strengths of the Yinsol resin cements or
Surface area 1660 sq. cm. g.
A No. 108 — Compressive strength at 28 days - - 3210
B Xo. 115-
C No. 112
Water - A - Plain
B - Yinsol resin
C - Dispersed
35 y 2 gals.
A comparison of a Vinsol resin cement with normal portland cement
and a dispersing agent from an economic point of view is even more
favorable for the latter than it would be if compared to undispersed
Portland cement or to portland-natural blends. If it is assumed that a
Vinsol resin cement will cost approximately the same as a similar
normal portland cement a rough comparison may be made. A loss in
strength, if the Vinsol resin is effective, of about 20 r 7 may be expected.
This is roughly equivalent to the strength imparted by one sack of
cement. Assume now that a mix of 4000 lbs. sq. in. is desired and that
a mix containing 5 sacks of normal portland cement will produce a
strength of 3000 lbs./sq. in. Now, by the addition of a dispersing agent
at a cost of approximately $0.30 per cu. yd. a strength of 4100 lbs. will
be secured. By the addition of one extra sack of portland cement at
a cost of $0.50 per cu. yd., a strength of 3900 lbs./sq. in. will be secured
(cf. Fig. VII Research Paper No. 36 for these strength data). With the
Vinsol resin cement which has lost 20% of the strength of the original
mix, that is, has reduced the strength of the 5 sack mix to 2400 lbs./sq.
in., in order to realize 4000 lbs./sq. in. it will be necessary to add
approximately 2 sacks of extra cement at a cost of $1.00 per cu. yd.
In other words, for a given strength, it will cost 20c more per cubic
yard with an undispersed normal portland cement compared with the
cement with a dispersing agent, and 70c more per cu. yd. with the
Vinsol resin cement (Table IX). There does not appear to be any
economic justification for the use of Vinsol resin cement.
Sacks Compressive Cost — Dollars Additional
Cement Strength (Cement plus Cost per
Type of Cement per Lbs. per Dispersing Cu. Yd.
Cu. Yd. Sq. In. Agent) Dollars
Normal portland cement 5 . 3000 2 . 50
Vinsol resin cement 5 . 2400 2 . 50
Normal cement dispersed 5.0 4100 2.80
To produce 4000 lb. concrete.
Normal portland cement 6.0 3900 3.00 0.20
Vinsol resin cement 7.0 4000 3.50 0.70
Normal cement dispersed 5 . 4100 2 . 80
Rosin, beef tallow, oil, and similar materials are very similar to
Vinsol resin. Like this last they also impair strength. On the whole
such information as is available would indicate that they are less
satisfactory than Vinsol resin and there is no evidence that they have
received any general acceptance although some experimental highway
installations have been made.
As far as the relation of these grinding aids to cement dispersion
is concerned it is the same as the relation between \ insol resin
ground cements and dispersion but, if anything, less favorable to
the grinding aids.
These cements can be dismissed with the statement that the same
properties, if desired, can be secured with Vinsol resin or a portland-
natural cement blend to better advantage. It follows that these prop-
produced more economically and without the
- loss in strength by use of a cemeni
ground with grinding aids ii
menl dispersing agent may bTgWnd
widely used grinding aid contamVaTa
Such cemente are in an entirely
ound with Vinsol resin beef tallow
produced and large quantities of
■-..•n.> in ih..,im|K-r ties f'.l
pear thai there are also^X* other
dingl I ,,,,1,,,^
\ ill. \\ sterproof < emeata
a smaT amount of some
probably ahTfanctta! to
is to Sake "the
me of cxceaa
" "■' ''••;-'.ii,i lo ii>rr<n-
/ ' /
/ ' /
| t /
i i f
• • en thai there an- available a number of special cements
velop some particular property of the cement
I e Mian would be the rase in a normal Portland cement.
either without substantial change in the other
■ expense of one or more of these other
erties winch it has been Bought to improve in this
ons, are early strength, corrosion (sul-
dration, capillarity, and workability in the
3S. This last property has also hem
ir and increased durability.
■ lalh applicable to the special cements as it
• i ' I i ame effect on them in reducing
consistency and making available a
1 < k>nsequently n has similar eff<
volume chai gth and
>mical means of producing a con-
to Lilar t\ pe of cement. Often n |
rt) without I lie attendant di
he relation i | dispcr
i 'i on in three w.-r
• re rum be
emei and a disp<
produced bj high <
i i Lent with respect
inability a- found in po/.xolan
• i round w nh grinding
.Hon alio holdj
Lant '1 \ pe \
dified low (,
' ■ • n : ■
al cement and a normal
i rid on t i i /) ;i r
•■■ " /- '" 'hi ituation foi the low heat cements 'J ype IVj
■n^M r<i? .• ;,/«' ' ;,-.,: i .,,. S |„. ( i;i , . 4 . f r if Jfi , ,
There is no conflict betwe
•ersing agent beai
economically and wi\ \\ re
to normal port land cement I
can be iecured al
dispersing agent ; in ot hei
and t he dispersii
appear de arable to u e i
car be realized ai minim i
of cement di pei ion
\i.i>ti r Builden Research Lab
c L \ i land, i ^ln» -
CEMENT DISPERSION PRINCIPLE
The principle of dispersion of cement is applicable to any type of
work involving cement in mortar or concrete. This work is of a very
varied nature and for different applications somewhat different proper-
ties are required. To accomplish these purposes most economical I \
the cement dispersing agent may be combined with other basic principles
for the improvement of specific properties of concrete and mortar, as
illustrated by the diagram below. These include pozzolanic activity,
stearate waterproofing, chemical hardening, and metallic aggregates.
The Master Builders Company has developed a group of products
adapted to various specific concrete and mortar applications. The
exclusive dispersing agent is incorporated in each of these products
in a manner to impart the maximum effect on the resultant structure
at minimum cost. These products are as follows:
Concrete (General) Pozzolith
High Early Strength Concrete High Early Pozzolith
Concrete (Exposed to Capillary
Moisture) Omicron Waterproofing
Floors — Heavy Duty ..Masterplate
Floors — Light Duty Mastermix
Colored Floors Colored Metalicron and Colormix
Brick Mortar Omicron Mortarproofing ("0. M.")
Colored Brick Mortar Colored Omicron Mortarproofing
Grouting and Maintenance Embeco
PRODUCT COMPOSITION DIAGRAM
THE MASTER BUILDERS COMPANY
Factories in Cleveland
Sales Offices in
All Principal Cities